Íslenskt mál og almenn málfræði - 01.01.2017, Page 132

summary
‘Poetry, phonology and poetic traditions’
Keywords: alliteration, phonemes, underlying representations, poetic traditions
Icelandic linguists have long debated the relationship between poetry and phonology. This
article responds to recent contributions to that discussion by Höskuldur Þráinsson (2014)
and Haukur Þorgeirsson (2013). Sections 2–3 discuss the putative role of underlying rep-
resentations in poetry. A proposed universal by Alexis Manaster-Ramer (1981, 1994) holds
that rules of versification may not be sensitive to representations deeper than phonemic.
This principle is tested here on Faroese poetry and found to hold.
Alliteration between words starting with orthographic h is discussed in section 4.
Höskuldur Þráinsson has argued that this alliteration is phonologically natural but since he
does not argue this based on underlying representations, as traditionally conceived, it is
not clear that the Manaster-Ramer universal is relevant – rendering the discussion in
Haukur Þorgeirsson 2013 irrelevant to Höskuldur’s argument. Höskuldur argues that
Icelandic children can perceive initial [ç] as /hj/ based on an innate aversion to voiceless
sonorants. This is an interesting theory that needs to be elaborated on for it to be testable.
What conditions need to hold for [ç] to be perceived as /hj/? Do speakers of (certain
dialects of) Norwegian and German also perceive word-initial [ç] as /hj/ or does Icelandic
have some further necessary properties to enable this perception?
Section 5 draws a distinction between two types of poetic traditions; lumping and
splitting traditions. In a lumping tradition, some distinction upheld in the language itself
is ignored for the purposes of poetry. It is proposed that this type of tradition is natural
and occurs relatively frequently; examples from early Icelandic poetry would include the
full rhyme of a and ǫ and alliteration between vowels and j. In a splitting tradition, a poet
upholds some distinction in his poetry which is not present in his or her normal speech.
Icelandic poets who make a distinction between words with initial hv and initial kv in allit-
eration but not in their everyday speech uphold a splitting tradition. This type of tradition
is artificial and difficult and requires knowledge of an archaizing orthography or other
learned materials. It is proposed that splitting traditions are rarer than lumping traditions.
Section 6 shows two examples of the role of orthography in poetry. In the poetry of
the 16th to 19th centuries, poets used the final inflectional [ɪ] to rhyme interchangeably
with stressed [ɪ] or [ɛ]. This was in accordance with the spelling of the time, where final
[ɪ] was written either as “i” or as “e”. But according to the Grammatica Islandica of Jón
Magnússon (1662–1738), the spelling with “e” was merely conventional and not in accor-
dance with the pronunciation, which seems to have been the same as today. The second
example is the loyalty that 20th century poets show to the standardized orthography in
rhyming unstressed syllables ending in n. Þórarinn Eldjárn carefully rhymes nýjungin with
skyn and iðnaðinn with sinn even though the spoken language has had no distinction
between the ending of nýjungin and iðnaðinn for centuries.
Section 7 presents two counter-arguments from reviewers of this article. One is that
alliteration is different from rhyme and even if rhyme can be shown to have a traditional
component this might not hold for alliteration. I argue that we would expect the weight of
tradition to be even more heavy in alliteration since Icelanders seem to learn rhyme effec-
Haukur Þorgeirsson132